Goods balanced loading system and loading method based on real-time modeling

Abstract

The invention relates to the technical field of cargo balanced loading, in particular to a cargo balanced loading system and method based on real-time modeling, and the system comprises the following modules: a sensing module, a control module, an algorithm module, a model database, a communication module and a display module, and the sensing module senses the loaded space of a carriage in real time through sensing equipment. The method has the advantages that real-time performance is high, real-time modeling is achieved, a stacking scheme is recommended in real time, residual space is calculated in real time, and existing stacking earnings and estimated stacking maximization earnings are calculated in real time; flexibility is improved, a scheme generated at a time is not depended on, stacking workers can freely stack, the optimal stacking scheme is calculated and recommended in real time, and the maximum income is obtained; a plurality of scenes of a loading and unloading point can be integrated, and the maximum transportation income is obtained in combination with an algorithm; the algorithm is wide in application scene, can be used for regular cubes and can also be suitable for cylindrical goods, conical goods and other special-shaped goods, and the use scene is not limited.

Description

Technical Field

[0001] The present invention relates to the technical field of balanced cargo loading, and in particular to a balanced cargo loading system and a loading method based on real-time modeling. Background Art

[0002] In the logistics, transportation and warehousing industries, packing operations are the key bridge connecting all links of the supply chain. In order to enhance the competitiveness of the industry and solve the packing problems that have long plagued enterprises, the packing solutions in existing technologies are based on process integration and core algorithm research, combined with mobile Internet information technology, and through visualization, to provide efficient, accurate and regulatory-compliant intelligent solutions for packing operations.

[0003] First, we conducted an in-depth analysis of the traditional packing process and identified the inefficient links and bottlenecks. Through process reorganization and standardized operations, we optimized the operation steps, reduced unnecessary waiting and repetitive work, and significantly improved the packing efficiency. At the same time, we introduced advanced packing algorithms that can calculate the best loading plan based on the size, weight, shape and stability requirements of the goods, ensuring the safe transportation of goods and maximizing the use of space.

[0004] In terms of technology application, we fully utilize the advantages of mobile Internet information technology. By developing a dedicated mobile application, operators can receive packing instructions in real time, monitor the packing progress, and make adjustments when necessary. This instant information exchange and feedback mechanism greatly shortens the response time and improves the flexibility and adaptability of operations.

[0005] In addition, some solutions also develop a set of 3D simulation software to make the packing process intuitive and easy to understand. Operators can use the software to preview the placement of goods, stacking effect and space utilization, so as to better understand the packing plan and make fine adjustments when necessary. This visual tool not only improves the accuracy of the operation, but also enhances the trust of customers.

[0006] Data-driven is another major feature of the existing solution. By collecting and analyzing relevant data on packing operations, such as packing time, cargo damage rate, customer satisfaction, etc., we can continuously optimize the process and improve service quality. Through data analysis, we can discover potential problems and room for improvement, and achieve continuous self-improvement.

[0007] In summary, the current containerization solution is a comprehensive innovation system that provides an efficient, intelligent and regulatory-compliant containerization business model for the logistics, transportation and warehousing industries through process optimization, algorithm application, mobile Internet technology and data-driven decision-making. The existing solution will help companies solve long-standing containerization problems, improve their competitiveness and win market favor.

[0008] However, the existing technology also has certain drawbacks and shortcomings. Most of the existing technologies currently generate packing plans in advance through data integration, while the actual production and transportation conditions are complex and changeable and cannot fully meet current needs.

[0009] The invention patent with application number CN202210493447.1 discloses a method, device, electronic device and storage medium for optimizing container space. It specifically discloses obtaining container information, information about goods to be loaded, and cargo stacking requirements corresponding to different goods in the cargo information to be loaded, and generating a stacking plan. The plan can generate a complete stacking plan through pre-input stacking requirements and cargo information. However, there may be temporary adjustments in the actual production stacking process, such as: cargo damage, cargo replacement, changes in transportation routes, adjustments to transportation requirements and other unpredictable risks. In addition, the plan is suitable for long-distance single packing or foreign trade transportation, and is not applicable to packing algorithms for small batches and multi-location loading and unloading. In addition, traditional packing algorithms are only applicable to packing of goods in approximately cubic shapes, while cylindrical, conical and other special-shaped goods are not currently supported. Therefore, the existing one-time generation of overall square stacking plans has the following disadvantages:

[0010] 1. Lack of flexibility:

[0011] Lack of flexibility: the solution obtained at the beginning of the assembly is different from the actual assembly process;

[0012] 2. Poor real-time performance:

[0013] The overall stacking plan generated at one time has poor real-time performance and cannot be updated in real time;

[0014] 3. Limited scope of application:

[0015] This method is suitable for scenarios where all loading and unloading points are unique, but is not applicable when the loading and unloading points are not unique.

[0016] 4. Limitations of algorithm use:

[0017] Packing algorithms are generally only applicable to packing goods that are approximately cubic in shape. Existing algorithms do not support conical or other irregular shaped goods. Summary of the invention

[0018] In order to overcome the technical defects in the prior art and effectively solve the technical problems in the background technology, the present invention provides the following technical solutions:

[0019] A cargo balanced loading system based on real-time modeling, comprising the following modules: a perception module, a control module, an algorithm module, a model database, a communication module and a display module;

[0020] The sensing module senses the loaded space in the carriage in real time through the sensing device; and senses the loaded weight in the carriage in real time through the pressure sensor;

[0021] The control module is used to process the perception data in real time, build a real-time model through the survey model library, and calculate in real time through the scheduling algorithm to give the overall cargo ratio plan and cargo placement schematic suggestions; during the stacking process, dynamically give the stacking suggestions and deviations for each cargo, and give the gap with the optimal solution;

[0022] The algorithm module is used to calculate the maximum profit and provide a loading suggestion plan based on the unit prices of different light and heavy cargo transportation when specifying the carriage and load capacity;

[0023] The model database includes a vehicle cargo box model and a transportation route model;

[0024] The communication module is used to communicate with the platform or field equipment;

[0025] The display module is used to display the real-time loading status.

[0026] As a preferred technical solution of the present invention, the vehicle cargo box model refers to the cargo box size, maximum load weight and nuclear load capacity of different types of vehicles.

[0027] As a preferred technical solution of the present invention, the transportation route model refers to the transportation distance, the unit price of light and heavy cargo transportation and the road weight limit, and can be presented in combination with a map.

[0028] As a preferred technical solution of the present invention, the communication module uses 4G or 5G for long-distance communication and uses WIFI or Bluetooth for near-field communication.

[0029] A loading method of a cargo balanced loading system based on real-time modeling comprises the following steps:

[0030] S1: Input the transport route and transport vehicle information, retrieve the model database; use the tool to regularly create and update the vehicle container model and transport route model;

[0031] S2: Input the loading cargo information and the geographical location of each type of cargo, call the algorithm module, dynamically calculate the recommended loading plan and corresponding benefits, and visualize the loading suggestions and transportation route restrictions;

[0032] S3: Load one by one according to the recommended loading route. During loading, the sensing module senses the current loading space and weight in real time, and loads and adjusts according to the real-time loading plan;

[0033] S4: According to the actual changes in the loading cargo ratio, follow up the final loading situation and provide the final loading plan and predicted revenue;

[0034] S5: Each actual transportation plan is stored in the database and used as a model training correction to continuously improve the accuracy of the model and algorithm.

[0035] Compared with the prior art, the present invention has the following beneficial effects: the present invention aims to solve the difficulties of the original technology by real-time modeling and real-time recommendation, and the specific advantages are as follows:

[0036] 1. High real-time performance:

[0037] Real-time modeling, real-time recommendation of stacking solutions, real-time calculation of remaining space, real-time calculation of existing stacking benefits and estimated maximum stacking benefits;

[0038] 2. Improved flexibility:

[0039] It does not rely on a one-time generated plan, and can allow stacking workers to stack as they wish. It can calculate and recommend the best stacking plan in real time to obtain the maximum benefit.

[0040] 3. Increased transportation revenue:

[0041] It can integrate multiple loading and unloading point scenarios and combine algorithms to obtain the maximum transportation benefits.

[0042] 4. The algorithm is applicable to a wide range of scenarios:

[0043] It can be used for regular cubes, as well as cylindrical, conical and other special-shaped goods, and its usage scenarios are not limited. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a framework diagram of the loading system of the present invention.

[0045] Figure 2 This is a schematic diagram of the sensing device deployment structure of the loading system of the present invention.

[0046] Figure 3 The figure is a flow chart of the loading method of the present invention.

[0047] Figure 4 This is a schematic diagram of real-time stacking plan recommendation and benefit display after two loading space updates of the present invention.

[0048] Figure 5 Flowchart for numerical simulation of object volume using Monte Carlo numerical simulation algorithm. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] See also Figure 1-4 The present invention provides a technical solution: a cargo balanced loading system based on real-time modeling:

[0051] System composition:

[0052] like Figure 1 As shown, the system includes the following modules: a perception module, a control module, an algorithm module, a model database, a communication module and a display module;

[0053] The perception module uses cameras, lidar and other perception devices to sense the loaded space in the car in real time; and uses pressure sensors to sense the loaded weight in the car in real time;

[0054] The control module is used to process the perception data in real time, build models in real time through the survey model library, and calculate in real time through the scheduling algorithm to give the overall cargo ratio plan and cargo placement suggestions; during the stacking process, it dynamically gives the stacking suggestions and deviations for each cargo, and gives the gap with the optimal solution;

[0055] The algorithm module is used to calculate the maximum profit and provide loading suggestions based on the unit prices of different light and heavy cargo transportation when specifying the carriage and load capacity;

[0056] The model database includes vehicle cargo box model and transportation route model. The vehicle cargo box model refers to the cargo box size, maximum load and nuclear load of different types of vehicles, and the transportation route model refers to the transportation distance, unit price of light and heavy goods transportation and road weight limit, and can be presented in combination with a map;

[0057] The communication module is used to communicate with the platform or on-site equipment. Specifically, 4G or 5G is used for remote communication, and WIFI or Bluetooth is used for near-field communication;

[0058] The display module is used to display the real-time loading status using media such as tablets and display platforms.

[0059] Deployment method:

[0060] 1. The algorithm platform and the perception devices included in the system are deployed on the same network. The perception devices send data to the algorithm platform in real time through the network for real-time modeling;

[0061] 2. Specific deployment (such as Figure 2 shown):

[0062] (1) Sensor deployment:

[0063] Spatial perception: At least 4 cameras are used to integrate perception and stitching to obtain a panoramic view and solve the problem of blind spots in perception. Specifically, the cameras include but are not limited to industrial cameras with the function of object contour recognition and uploading point cloud to facilitate real-time modeling.

[0064] Weight perception: At least 4 pressure sensors are distributed on both sides, front and back of the car for balanced perception.

[0065] (2) System deployment:

[0066] Contains 1 platform and 2 types of sensors.

[0067] Workflow:

[0068] 1: Input the transport route and transport vehicle information, and retrieve the model database; use the tool to regularly create and update the vehicle container model and transport route model;

[0069] 2: Input the loading cargo information (multiple categories are optional) and the geographical location of each type of cargo, call the algorithm module, dynamically calculate the recommended loading plan and corresponding benefits, and visualize the loading suggestions (cargo ratio, loading sequence, loading route) and transportation route restrictions (map display, such as height limit, weight limit, where height limit affects loading sequence), etc.;

[0070] 3: Load one by one according to the recommended loading route. During loading, the perception module senses the current loading space and weight in real time, and loads and adjusts according to the real-time loading plan. Among them, the perception algorithm needs to be calibrated, and the sensor calibration is performed for goods of different volumes, shapes, and weights placed in different compartments;

[0071] 4: According to the actual changes in the loading cargo ratio, follow up the final loading situation and provide the final loading plan and predicted revenue;

[0072] 5: Store each actual transportation plan in the database and use it as a model training correction to continuously improve the accuracy of the model and algorithm.

[0073] Specifically, the specific workflow of this embodiment is as follows Figure 3 As shown:

[0074] 1. After starting, input the transport route, transport vehicle information, and the loading and unloading cargo volume information of multiple loading and unloading points, retrieve the model database (for example: starting point a, end point d, passing points b and c, cargo box capacity is v, cargo capacity from a to d is 30%*v, cargo capacity from a to b is 10%*v, cargo capacity from b to d is 70%*v, cargo capacity from c to d is 10%*v), and generate an optimal solution with the maximum transport revenue through algorithm calculation;

[0075] 2. Drive to the next transportation point and determine whether it is the last transportation point for loading and unloading goods;

[0076] (1) Yes, it is completed;

[0077] (2) No, proceed to the next process;

[0078] 3. The cargo loading and unloading recognition model identifies the cargo handling direction and determines whether it is loading or unloading;

[0079] (1) Loading: Determine whether the object is a standard part based on its shape (such as sphere, cuboid, cylinder, etc.);

[0080] For standard goods (such as spheres, cuboids and cylinders, etc.), directly use the existing volume calculation formulas for spheres, cuboids and cylinders, and perform approximate volume calculations based on the calibrated length, width, height or radius to obtain the volume Vn0, and subtract β*Vn0 from the total remaining volume V0 to obtain the real-time remaining volume (where α and β are constant coefficients); for non-standard goods, use visual equipment for real-time modeling, and call the algorithm module to calculate the volume Vn1 of the object through Monte Carlo simulation (numerical simulation method), and subtract α*Vn1 from the total remaining volume V1 to obtain the real-time remaining volume;

[0081] (2) Unloading: Determine whether the object is a standard part based on the shape of the object (such as sphere, cuboid, cylinder, etc.);

[0082] For non-standard goods, visual equipment is used for real-time modeling, and the algorithm module is called to calculate the volume Vn2 of the object through Monte Carlo simulation (numerical simulation method), and the total remaining volume V2 plus a*Vn2 is used to obtain the real-time remaining volume; for standard goods, the existing volume calculation formulas for cuboids, spheres, and cylinders are directly used to approximate the volume according to the calibrated length, width, height or radius to obtain the volume Vn3, and the total remaining volume V3 plus β*Vn3 is used to obtain the real-time remaining volume; for example: regular cuboids, cylinders, etc. are directly obtained using the existing physical volume calculation formula (where α and β are constant coefficients);

[0083] 4. When the changes in the space inside the vehicle are perceived in real time and modeled in real time, the algorithm module is requested to exhaustively list the loading solutions and recommend the optimal loading and distribution ratio based on the principle of maximizing transportation revenue;

[0084] 5. Regenerate the comparison between the real-time loading plan and the estimated loading plan, and give loading suggestions;

[0085] 6. Determine whether loading and unloading has been completed;

[0086] (1) Yes, it is completed;

[0087] (2) No, return to workflow 3 and continue the workflow.

[0088] Specifically, the use process of the Monte Carlo simulation algorithm in the above workflow 3 is as follows (eg Figure 5 shown):

[0089] (1) Obtain a three-dimensional model of an irregular object through modeling;

[0090] (2) Use simple geometric shapes of regular known volume (such as cubes and spheres) to enclose irregular objects;

[0091] (3) Generate a large number of points within the enclosed object, the positions of which are randomly determined according to a uniform distribution or other appropriate probability distribution;

[0092] (4) For each generated random point, use algorithms and mathematical formulas to determine whether the point is located inside the irregular object;

[0093] (5) Count the proportion of all random points that fall inside irregular objects;

[0094] (6) Assuming that the volume of the bounding box is V_box, and the proportion of the number of points falling within the irregular object to the total number of points is p, the volume of the irregular object V_object can be estimated by the following formula: V_object = p*V_box.

[0095] In addition, if Figure 4 As shown, the present invention shows a real-time stacking plan recommendation and benefit example after two loading space updates, which is similar to the recommendation algorithm of existing mature applications. It recommends loading plans in real time, compares the benefits with the actual loading plans, and displays them graphically to guide on-site staff in stacking, but does not affect the actual stacking of the staff. It is for reference only and is not a mandatory stacking requirement.

[0096] The workflow of the present invention forms an optimal loading plan by combining the distance and sequence of the loading and unloading points, and reserves space for the next loading and unloading point in advance to maximize space utilization.

[0097] When using:

[0098] For specified cargo box size and load restrictions, real-time modeling of the cargo in the cargo box and combined with the built-in weight balancing algorithm can achieve optimal configuration of volume and weight of both light and heavy cargo (different densities, such as large household appliances with low density and small steel coils with high density) in the cargo box to obtain maximum profit for a single transport.

[0099] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A cargo balancing loading system based on real-time modeling, characterized by: It includes the following modules: perception module, control module, algorithm module, model database, communication module and display module; The sensing module senses the loaded space in the carriage in real time through the sensing device; and senses the loaded weight in the carriage in real time through the pressure sensor; The control module is used to process the perception data in real time, build models in real time through the survey model library, and calculate in real time through the scheduling algorithm to give the overall cargo ratio plan and cargo placement schematic suggestions; during the stacking process, dynamically give the stacking suggestions and deviations for each cargo, and give the gap with the optimal solution; The algorithm module is used to calculate the maximum profit and provide a loading suggestion plan based on the unit prices of different light and heavy cargo transportation when specifying the carriage and load capacity; The model database includes a vehicle cargo box model and a transportation route model; The communication module is used to communicate with the platform or field equipment; The display module is used to display the real-time loading status.

2. The cargo balanced loading system based on real-time modeling according to claim 1, characterized in that: The vehicle cargo box model refers to the cargo box size, maximum load weight and nuclear load capacity of different types of vehicles.

3. The cargo balanced loading system based on real-time modeling according to claim 1, characterized in that: The transport route model refers to the transport distance, unit price of light and heavy cargo transport and road weight limit, and can be presented in combination with a map.

4. The cargo balanced loading system based on real-time modeling according to claim 1 is characterized in that: The communication module uses 4G or 5G for long-distance communication and uses WIFI or Bluetooth for near-field communication.

5. The loading method of the cargo balanced loading system based on real-time modeling according to any one of claims 1 to 4, characterized in that: The following steps are involved: S1: Input the transport route and transport vehicle information, retrieve the model database; use the tool to regularly create and update the vehicle container model and transport route model; S2: Input the loading cargo information and the geographical location of each type of cargo, call the algorithm module, dynamically calculate the recommended loading plan and corresponding benefits, and visualize the loading suggestions and transportation route restrictions; S3: Load one by one according to the recommended loading route. During loading, the sensing module senses the current loading space and weight in real time, and loads and adjusts according to the real-time loading plan; S4: According to the actual changes in the loading cargo ratio, follow up the final loading situation and provide the final loading plan and predicted revenue; S5: Store each actual transportation plan in the database and use it as a model training correction to continuously improve the accuracy of the model and algorithm.

Images